Adults over fifty years old are twice as likely to have tinnitus and four times as likely to be disabled by it than those under fifty. While high- frequency sensorineural hearing loss predicts best who will have tinnitus, the presence of depression--not properties of the tinnitus itself--appears to predict best who will become disabled by it. This study will examine the relationship between depression and tinnitus disability through a randomized, placebo-controlled, double-blind trial of nortriptyline in two groups of disabled tinnitus patients: those who meet DSM-III-R criteria for Major Depression and those who do not. It will thereby be possible to determine if tricyclic antidepressant treatment is effective for the reduction of tinnitus disability under these rigorous conditions and if the effectiveness of tricyclic treatment is due to its effects on depression or to a more direct effect on tinnitus itself (such as the analgesic effect of tricyclics in chronic pain syndromes). Impact of treatment upon tinnitus severity will be assessed in three areas: physiological severity (by changes in audiometric measures of tinnitus). Functional severity (by patient self-report of changes in tinnitus-related disability), and burden of illness (by spouse interview and self-report). This detailed analysis of the interaction between depression and tinnitus should provide greater definition of the psychosocial parameters that mediate the production of disability by medical illness in the elderly. By providing a clearer understanding of the relation of symptoms to disability, we can significantly advance our general understanding of medical/psychiatric co- morbidity in the context of aging.